Re: Arabic letters separated by markup

At 10:29 AM 6/16/2005, Steve Zilles wrote:
>For what it is worth the following text comes from the XSL 1.0 REC 
>concerning when a ligature substitution is to be done. From section 4.7.2 
>Line Building:
>
>...substitutions may occur because of addition of hyphens or spelling 
>changes due to hyphenation, or glyph image construction from syllabification,

what is "glyph image construction from syllabification" ? Indic conjuncts? 
Conjoining Jamos?

>  or ligature formation.
>
>Substitutions that replace a sequence of glyph-areas with a single 
>glyph-area should only occur when the margin, border, and padding in the 
>inline-progression-direction (start- and end-), baseline-shift, and 
>letter-spacing values are zero, treat-as-word-space is false, and the 
>values of all other relevant traits match (i.e., alignment-adjust, 
>alignment-baseline, color trait, background traits, 
>dominant-baseline-identifier, font traits, text-depth, text-altitude, 
>glyph-orientation-horizontal, glyph-orientation-vertical, line-height, 
>lineheight-shift-adjustment, text-decoration, text-shadow).
>
>This indicates a bias to honoring the author's/user's styling choices over 
>ligature formation. I am not sure how well these paragraphs have been 
>tested in practice.

That bias seems correct where the typographical effects are also merely 
stylistic. However, without knowing more about the rationale I am very 
skeptical when it comes to overriding orthographically required effects, 
such as required ligatures, as well as hyphenation-related spelling 
changes. In these cases it would be preferable to have an explicit 
parameter that requests the suspension of these orthographic rules for a 
special purpose.

In this context, if we assume that authors are using a wysiwyg editor, they 
might be alerted to unusual behavior around required ligatures, however, 
they would most likely not even realize that their action also affects the 
hyphenation behavior, as that is exhibited only when a line break actually 
involves a specific hyphenation. Where styles are applied to documents are 
created from application of a template to e.g. an XML document, in other 
words, where no-one actually edits the final form, this preference for 
style over orthography seems troublesome.

However, in the course of this discussion, many people have pointed out the 
difficulties in attempting to process a ligature (or any of these other 
substitutions) when the styles across it are incompatible. I quite agree 
and don't want to be understood as trying to argue that rendering engines 
should suddenly do the impossible. However, fantasai's 5 points strike a 
different balance in terms of which features to preserve in a fallback, and 
on the whole, they seem preferable to the style over substance stance that 
Steve reports from CSS.

A./